The current thesis aims to explore the potential of producing large-scale structural cast glass elements with the help of 3d printed sand moulds. Until now, mostly small blocks from structural cast glass have been generated in the building industry, revealing a research gap regarding massive components. Apart from the size, the unlimited shape potentials of cast glass have hardly been explored so far. This is because glass casting can be a very challenging process when it comes to the solidification of the material which needs to be completed under specific program settings. Except from the brittleness of the body during the annealing process, it can take months, even years for a large-scale element to be completed, because of the full controllable conditions under which the annealing process takes place. In addition, the mould types that are currently used for cast glass components present several restrictions such as the manufacturing costs, the accuracy or the sizing. An existing glass slab in the Acropolis museum, in Athens, was chosen to embody the research goals. The slab is located on top of archeological discoveries making the need for transparency necessary. The current slab consists of float glass panes with a concrete substructure which severally restrains the viewing. The thesis aim is to develop a free-form monolithic slab exclusively from cast glass without the need of external substructure. In order to address the above demands, the topology optimization method was introduced. Several algorithms of topology optimization have been developed. A compliance-based optimization was demonstrated as the most appropriate for the specific project. Obtained from the literature framework, the main principles that the design follows gravitate towards the structural, cast glass and manufacturing requirements. The external applied load, in addition to the allowable structural values were calculated. The importance for a flat upper surface, where a safety float glass layer of a permitted deformation should be laminated, was established. Regarding the cast glass demands, it is necessary to acquire a design with smooth surfaces, round edges and uniform distributed material throughout the entire model. For the current project, the optimized slab is decided to mainly consist of borosilicate glass.. A slab separation is crucial for transportation reasons. After the definition of the main design principles that the final product should respect, the design development was conducted with the combination of three methods: topology optimization, manual design and structural validation. The final result of the optimized slab successfully fulfills the preset design criteria, as the obtained structure presents 55.2% less mass than the solid version that was initially designed to replace the existing slab in the museum. In addition, a more efficient material distribution in the optimized model renders the slab more compatible for the cast glass annealing process. The result of the topology optimization technique was a quite complex geometry which is not that effective to be produced in the conventional moulds that are currently available for cast glass applications. Therefore, a new technique needed to be explored, which follows the 3d printing evolution in the building industry. The main restriction of the printed moulds is the limited available sizing that the commercial printers provide. As a solution to this problem, the separation of the large-scale mould into smaller segments was studied, as well as the development of the connecting system of these parts. The entire fabrication technique, from the moulds printing method, to the assembly components and glass melting was investigated. Finally, the transportation and integration of the slab into the building was explored, concluding with structural solutions that respect the museum and the assembly order that the slab can be attached on it.